ABSTRACT:
Two sets of papers investigating the top-of-the-mind problems of today open this Journal of Management Information Systems (JMIS) issue that completes the 40th-anniversary volume of the Journal. The first set presents the results of the impactful research on the interaction between humans and artificial intelligence (AI). The current and coming generative AI generates optimism regarding the new levels of human accomplishment when assisted across their working and private lives. Its promise contains a quantum productivity growth, which can lift billions out of poverty and toward expanded horizons. More specific to the information-systems’ (IS) concerns, the ability of the generative AI to generate software code leads directly to what we may call implicit development by the end users who are able to specify their needs in a natural language. As a general-purpose technology, the new AI also produces high levels of anxiety regarding the availability of employment and, beyond that, the very survival of the humans when facing AI. The work on the guardrails is clearly to be praised, if its effectiveness in the present geopolitical environment is to be questioned. The two papers opening the issue address, respectively, the role of AI in the competition and in the collaboration with people.
The second set of papers deals with cybersecurity, which has become a C-suite level corporate concern and to which the research published in the JMIS over the recent decades has contributed weightily. With the security of infrastructures and corporate systems increasingly endangered by the ever more sophisticated attacks by hostile actors, the progressing and deepening reliance on IS heralds the ever-greater exposure to the attacks of various levels of persistence. It is to be noted that the promise of generative AI also contains the promise of greater vulnerability to cyberattacks. This is blindingly obvious if you consider only the availability of the large language model systems in the open-source software space. Therefore, the two problems—that of salutary deployment of AI and that of cybersecurity are—interrelated. Our research needs to surface a broader understanding of cybersecurity, confronting the AI-generated threats. Equally, we need to learn how to use powerful AI-based systems to confront these threats. A new and expanded cybersecurity research agenda is needed.
A great variety of tasks are becoming a contested ground when performed by AI rather than humans. We have to keep in mind while saying this that all these tasks are to be performed ultimately for the human benefit. We also need to remember and research intensively (bracketing out of the present narrative), the employment effects of the AI deployment. The managerial task of employee performance evaluation would rank among the more controversial assignments given to AI, with its acceptance raising doubts. The comparative fairness of human managers versus AI is studied in a field experiment by the authors of the opening paper. Shaojun (Marco) Qin, Nan Jia, Xueming Luo, Chengcheng Liao, and Ziyao Huang conducted comparative data-intensive performance evaluations in a fintech company to obtain causal results. The work shows that the evaluated employees considered AI to be both fairer and more accurate than managers. Moreover, the perceived deficiency in the fairness of the evaluation by managers was the lens through which the employees saw the (in)accuracy of the human evaluation. In other words, to compete with AI, the managers need to be more human in their relationship with the employees. The authors’ theorizing and empirical results allowed them to offer a nuanced discussion of how human managers can compete with AI in the performance-evaluation task. The work is generative, as it will expand further research into the effectiveness of AI in managerial tasks.
The collaboration between AI and humans on prediction tasks is studied empirically by Elena Revilla, Maria Jesus Saenz, Matthias Seifter, and Ye Ma. Employing, as does the preceding paper, a field experiment (here, in the retail industry), the authors assess the effects of the human collaboration with AI in demand forecasting. They trichotomize the human-AI collaboration into three different degrees of human interventions and study their differential effectiveness in different forecasting environments. The environments are further characterized by their uncertainty and forecasting horizon. The paper culminates in a framework that contributes significantly to our understanding of the potential role of humans in the AI-driven forecasting and the results we may expect.
As noted previously, with the vast and rapidly increasing role of information systems (IS) in the functioning of our organizations and our lives in general, and with the increasingly organized attack-ready agents, the importance of cybersecurity cannot be overstated. Our research in the field, much of it published in the JMIS¸ has offered multipronged contributions, from more technologically-oriented to the behaviorally-inclined. Obviously, combinations of various approaches are needed in a layered protection environment. Here, Paul Benjamin Lowry, Gregory D. Moody, Srikanth Parameswaran, and Nicholas Brown present a highly comprehensive study of the effectiveness of fear appeals in the shaping of the employee cybersecurity posture. Using an innovative methodology and grounding themselves in several well-exercised theories, the authors meta-analyzed numerous studies to demonstrate the central role of fear and show the need to combine fear control with danger control for behavioral cybersecurity measures.
The risks to cybersecurity of a firm depend on multiple factors. The next paper of the issue brings to the forefront the role of the technological innovativeness of a focus firm. Qian Wang, Eric W. T. Ngai, Daniel Pienta, and Jason Bennett Thatcher empirically investigate the association between the IT innovativeness and the risk of data breaches. The authors base their research in the organizational learning theory and offer a longitudinal study that leads them to their conclusions: When we innovate technologically, we incur a greater risk of data breaches. Beyond that, the data-breach risk depends on the complexity and dynamism of the firm’s operating environment. The conclusions are clear and important. Of course, we want and need to innovate. However, when we do so, we also have to provide increased funds and structures for battening our hatches. The data the authors deployed show that the firms had not done so.
The development of the metaverse—a virtual social world of work, play, and broader life—is progressing, largely in a bottom-up manner. Avatars representing individuals are communicating here as the mediators of the would-be traditional human-to-human communication. Here, Ching-I Teng, Alan R. Dennis, and Alexander S. Dennis argue that this communication mode brings changes to the users’ (now represented by their avatars) identification with the community where they act and to their loyalty to the community. The model the authors develop and test allows them to offer theoretical contributions to alter our received ideas of social presence and community identification. The contributions to the design of the serious and less-serious games and to the fostering of metaverse are also clear.
With the next paper, Franck Soh and Varun Grover help us better understand our world of the vastly modifying supply chains underpinned by information technology (IT) and, in fact, being presently re-built to deliver this technology. Mobile apps are, at this point, a weighty component of the technological environment, just as some of the more prominent of them are an object of international controversy. The authors of the present work define a “foreignness liability” as the disadvantage that a foreign IT complementor, such as a mobile app startup, faces when compared to a domestic startup. The authors developed and tested a model of the performance of such startups in a variety of circumstances, including the varying cultural distance between the country of origin and the country of investors, and the dynamism of the environment. Are the app platforms offering a level playing field? The answer is more nuanced than one would originally think and that is why this work is of value. Beyond that, the work breaks the ground for a broader research on the building of IT supply chains.
As we have by now fully realized, cryptocurrencies are here to stay. We also know that we need to know much more to begin to understand—and at some point control—these marketplaces. The paper by Mariia Petryk, Liangfei Qiu, and Praveen Pathak makes a contribution here by exploring the impact an open-source software community can have on the price of a cryptocurrency. In their research, the authors build on the dual nature of the cryptos as a financial asset and a software program maintained in the open-source software (OSS) sphere. Relying on the signaling theory and econometrics, the authors are able to use the events in the communitarian software development (such as forks) as predictors of the crypto price movements. The innovative research idea clearly opens many avenues to further exploration and exploitation.
Online reviews have become a major component of both online and offline commerce, and a major subfield of IS research. One area that has been rather under-researched is social aspect of review influence. Are the reviews by more socially engaged users more influential? Are the more socially engaged users more active in reviewing? How much reciprocity is there between the reviewer and a follower in voting for the review? These are the questions answered empirically by Yinan Yu, Warut Khern-am-nuai, Alain Pinsonneault, and Zaiyan Wei in the next paper. Using an important dataset from a major review platform, they establish that the helpfulness scores of reviews are heavily influenced by the social relationships of the reviewers—and thus cannot be taken as fully objective value assessment of the reviews. Caveat lector.
To continue the theme of online commerce, the next paper turns to the online reward programs in the omnichannel retail environment of today. Yoonseock Son and Wonseok Oh deploy valuable matched datasets reflecting the reward-redemption behavior along with the transactional data to provide a discriminating study of what they call “digitalization of loyalty.” Monitoring the reward-redemption pattern enables the offeror firms to gauge the evolution of customer engagement and relationship with the vendor. The results obtained by the authors show that using the reward app increases customer engagement, yet the volatility of this engagement points to a higher probability of customer churn. We can also see that the vendors can learn much more from studying the reward-redemption patterns in order to satisfy and keep their customers.
The concluding paper of the issue delves further into the investigation of the brands’ competition in the online marketplaces. Meihua Zuo, Spyros Angelopoulos, Carol Xiaojuan Ou, Hongwei Liu, and Zhouyang Liang provide an optimization model for the brand offerings in these markets. With the ever- shorter product-development cycles, the firms face the challenges of the external competition as well as—and in particular—the intra-brand cannibalization of their own products. The authors argue, therefore, that the more traditional methods of product offerings need to be replaced by the models that take these developments into account and consider the intra-brand cannibalization as a highly significant factor. In order to do so, Zuo and colleagues deploy network theory to arrive at highly original recommendations, well based in the theory the authors muster and the empirics they bring to bear on their theorizing.